Wang, J., X. Li, L. Lu, and F. Fang. 2013. Estimating near future regional corn yields by integrating multi-source observations into a crop growth model. European Journal of Agronomy 49:126-140.
|Title||Estimating near future regional corn yields by integrating multi-source observations into a crop growth model|
European Journal of Agronomy
Abstract Regional crop yield estimations play important roles in the food security of a society. Crop growth models can simulate the crop growth process and predict crop yields, but significant uncertainties can be derived from the input data, model parameters and model structure, especially when applied at the regional scale. Abundant observational information provides the relative true value of surface conditions, and this information includes those areal data from remote sensors and ground observations. Data fusion technology integrates the advantages of crop growth models and multi-source observations, and it provides an innovative means for making precise regional corn yield estimations. A regional corn yield estimation framework based on two types of observation-model fusion methods is recounted in this paper. First, a 2008 application of the WOrld FOod Studies (WOFOST) growth model to the Yingke Oasis of Gansu province in northwest China suggested this method of simulating corn growth trends and yields, with attention to carbon absorption in particular. Second, this study applied a simulated annealing algorithm to obtain an optimized vector of parameters for the WOFOST model by using local multi-source data. After parameter estimation, the root mean square error (RMSE) of the simulated yield decreased from 1676.00 kg ha−1 to 4.00 kg ha−1. Moreover, the correlation coefficients between simulated and observed gross primary production (GPP) from 2009 to 2011 were 0.941, 0.967 and 0.962. Validation showed that a parameter estimation algorithm can reduce parameter uncertainties. Afterwards, the optimized model was used in a sequence data assimilation algorithm together with regional CHRIS leaf area index (LAI) data to incorporate spatial heterogeneity and evaluate model performance in estimating the near future regional corn yields. The general crop growth curve and final yield prediction were adjusted by using a real-time LAI variable update of the WOFOST model in each simulation unit. Numerical experiments on the sequence filter showed that the assimilation process can provide accurate regional estimations of crop growth and final yield on the basis of yield statistics from 50 sample points. The RMSE of the regional yield estimation at 50 sample points was 339.14 kg ha−1. Finally, by fusing a whole CHRIS-LAI image over the corn planting region of Yingke Oasis, a precise spatial distribution map of the estimated corn yield was obtained.
|This literature is not included PDF(How to submit?)|
Contact SupportNorthwest Institute of Eco-Environment and Resources, CAS 0931-4967287 email@example.com
LinksNational Tibetan Plateau Data Center